Ted’s Rants and Raves by Ted M. Young

I generally don’t have much to say about TechCrunch. I read it pretty often to see where people are putting their attention (or at least what TechCrunch thinks is the case) since I’m not really in that audience — I don’t visit Facebook (let alone have a profile), I don’t Twitter, and I use my cell phone for making phone calls and the occasional (once a month?) text message. However, a recent post on TechCrunch ticked me off because of the way it plays with statistics. The post is here and it claims that GMail will catch Yahoo! Mail, in terms of unique visitors, sometime in 2010. In and of itself, such predictions are not a problem, the problem is the way this prediction was arrived at: current percentage growth rate from November 2006 to November 2007. That’s right, a single number over an arbitrary time period is used as the sole basis for the prediction, and I quote: “Yahoo showed 3.21% growth for the 12 months to November 2007 compared with Gmail’s 53.60%”. Even worse, this single statistic is used to create Chartjunk:

From http://www.techcrunch.com/2007/12/25/2007-in-numbers-more-people-using-yahoo-mail-this-christmas-than-gmail/

Why the 12-month time-period? Why not compare the growth since Gmail started? Or use projections from the same time-periods for each service, i.e., compare the first two years of Yahoo Mail to the first two years of Gmail? Or pretty much any kind of deeper analysis that takes more than 5 minutes of thought? Does anyone really think that Gmail will continue to grow over 50% every year for the next 3 years? Or that Yahoo! Mail will only grow a few percent every year for the next three years? What makes the past year so special?

Not to mention, does anyone really care? I think the real issue is how much money are the two companies making? And what about the OEM-like agreements that Google has been making, i.e., becoming the email service provider for universities, companies, etc.?

Ugh. I hate Chartjunk and the simplistic statistics that back them up.

MarketWatch recently released a major upgrade to their portfolio tracking web application. Based on some reviews I’ve read elsewhere (including the one at TechCrunch), I decided that I should take a look. I currently use Google Finance’s portfolio for tracking my stocks because it made it easy to import my list of stocks by pasting in a CSV (comma-separated values) file into an input box. Unfortunately it has some glitches, such as not reporting the price of Nintendo correctly (though it uses the correct price for calculating the actual value of my shares), and no customization, and no charting.

Here’s Google’s quote for Nintendo:

Notice that the price (which should be the closing price of 76.10, since this was quoted well after the market closed) is the same as the opening price, 74.90, which is wrong. The fact that this is an ADR (American Depository Receipt) quoted on the “pink sheets” is no excuse:

And here’s Yahoo’s quote for Nintendo (taken within minutes of the Google snapshot)

Sure, Yahoo Finance could use a bit of a facelift in terms of layout, but at least they get the quotes right:

So, Yahoo gets the quotes right, but their portfolio system is really old — I don’t think it’s changed that much in the past 5 years (if not longer). Needless to say, it offers no importing of information, so I while I use it as my main quote source (delayed, of course, like all free quote sources), and for the news links (much more comprehensive than Google, if not overly so), it doesn’t serve my portfolio tracking needs. And, as I mentioned already, Google’s portfolio doesn’t have much in the way of features and is underwhelming in the news links area, which is especially surprising since they could just pull in the feeds from Google News.

Not Moving To MarketWatch’s Portfolio Tracker

When I first went to the new MarketWatch “My Portfolio”, it asked me if I wanted to import my old portfolio. I wasn’t quite sure what it meant, but I figured I’d say yes and see if it provided some way to import a portfolio from other web sites or at least from a file like Google Finance does. Alas, no: it was only for importing a portfolio that I might have already had on MarketWatch. Since I didn’t have one — though I wasn’t sure at first — it couldn’t import anything. I then looked around and couldn’t find any option for importing an existing portfolio. Since I wasn’t about to re-enter the information contained in my portfolio: number of shares, purchase price and date, commission, etc., I pretty much stopped there. Because there was no sample portfolio (which would have shown off its features), I entered a few stocks just to see how it looked (and I’ll have a closer look at the user interface soon), but there’s nothing I hate more than having to duplicate work that I’ve already done. And this could have been so easily avoided by providing a way to upload a spreadsheet (in either CSV, Excel, or OpenOffice format — or even all three!) or a pasted HTML table, i.e., one that I could have copied from another site. A fancy AJAX-based data-oriented web application doesn’t help if you haven’t thought about how a user is going to get their data into your application.

Oh well. Though it does open the door for a web-based service that could move your portfolio data around from site to site or even from, say, Quicken to Google Finance. Hmmm.